Ten years after the start of transition, there are many puzzles we still have to live with. Why did all countries experience strong declines in output at the outset of economic transformations and most of them are slowly, if at all, recovering from this "transitional recession"? How can these L-shaped patterns of GDP be reconciled with a shift from a less efficient to a more efficient economic system? Why were (and still are) unemployment pools of these countries so desperately stagnant in spite of the radical transformations going on? Why was unemployment dynamics so much different between, on the one hand, the Czech Republic, and, on the other hand, the other members of the Visegrad group? Why were employment-to-output elasticities negligible in Russia compared not only with Western countries, but also with the countries now knocking the door of the European Union? In this paper it is argued that many of these puzzles can be explained by simply taking on board labour supply. Surprisingly enough, the literature on the economics of transition has devoted little, if any, attention to labour force participation decisions. In the models of the optimal speed of transition (OST) literature, the labour force is generally assumed to be fixed. All the action takes place on the demand side. No mention is made to labour supply factors, the unsustainability of full employment at low wages in the absence of coercive power and the role played by non-employment benefits in inducing large flows to inactivity. The model developed in this paper allows for labour supply to play a key role in the transition by introducing three basic mechanisms in the Harris-Todaro type of models of the OST literature. First, room is made for frictions in the shift of workers from the old to the new sector. Second, job-to-job shifts are not ruled out: employers are free to choose their recruitment pool, that is, whether to hire from the unemployment ranks or among the employees of the old sector. Third, those without a job are allowed to make a non-trivial decision between searching or not searching a job. The model generates locking-in effects at the micro-level, and unemployment persistence at the aggregate level. The initial steps of transition are crucial in determining the importance of these locking-in effects. When the initial market-oriented reforms promote large flows from the old-sector to inactivity, it is very likely that employers in the new sector will be reluctant to hire from the ranks of the unemployed, as many of those without a job are not actually seeking. Low job finding probabilities in turn induce "discouraged worker" effects thereby those without a job do not actively seek a job, as their outside opportunity looks more appealing than spending a long time in job search efforts, having a very low chance to succeed. The model has important policy implications. Among these, it suggests that the emphasis placed by the OST literature on measures winning the resistance of insiders to restructuring, e.g., "buying-off" workers in the old sector, is ill-placed and possibly conducive to wrong policy prescriptions. By putting in place at the outset overly-generous non-employment benefit schemes, conditions were created for having stagnant unemployment pools throughout the transition. Long-duration unemployment made these promises unsustainable because generous non-employment benefits had been conceived for unemployment of a shorter-duration. Moreover, this tightening did not significantly reduce the duration of unemployment. Rather than starting with generous non-employment benefits and then subsequently cutting them down, the right sequence should have been the other way round.